Financial services firms eye machine learning

Machine learning is transforming the way financial services operates, as a growing number of Australian firms are adopting it to make sense of big data and to improve productivity.

According to Bloomberg, more than half (55%) of Australian financial services companies are investing in adopting or applying machine learning to better business processes and operations. One-in-five (19%) have already embarked on using machine learning solutions.

Based on its survey, two areas are fuelling greater machine learning activity in financial services: the increasing complexity of data interaction; and the rise of quantitative investing.

Bloomberg global head of machine Gideon Mann defines machine learning as software that incorporates "fairly complicated statistical models" by using optimisation techniques to fit certain parameters.

New York-based Mann, who was recently in Sydney, told Financial Standard that different organisations are at different levels of maturity in how they're approaching machine learning.

Already machine learning is being applied to digest unstructured data emanating from images, voice and multi-lingual sources. Mann added that finance professionals are increasingly looking at ways to integrate unstructured data into an overall model.

The most advanced hedge funds or quant hedge funds are currently gaining value out of machine learning, while others are "running really fast to catch up," he said.

Mann predicts greater application of machine learning will be evident in end-to-end strategy development, deep learning for time series analysis, and analysing data across languages.

In Australia, 46% of respondents in the Bloomberg survey are using machine learning to solve sophisticated intelligence when trading, while 34% use it to process complex data in a better way.

One-in-five (22%) said the ability to automate workflows is propelling interest in machine learning.

However, there is still not enough expertise in the area, according to 33% of respondents. Two other major drawbacks is the insufficiency of quality data, as well as having to carry out multiple, complex operations associated with it.

Richard Lane, the head of digital security of money transfer company OFX, said machine learning is a critical area to keep an eye on.

The deployment of machine learning will be a "huge area of development" over the next one to two years, particularly in cyber-security defence, Lane told a recent BAE Systems and Australian British Chamber of Commerce event.

Australia's financial services industry is a high-profile target internationally for cyber-security threats, but the good news is that firms are proactive in adopting security technology, he said.

Read more: BloombergFinancial StandardGideon MannOFXRichard Lane
Link to something QLjSY2P8